Antigenic evolution of SARS-CoV-2 in immunocompromised hosts

Abstract

Prolonged infections of immunocompromised individuals have been proposed as a crucial source of new variants of SARS-CoV-2 during the COVID-19 pandemic. In principle, sustained within-host antigenic evolution in immunocompromised hosts could allow novel immune escape variants to emerge more rapidly, but little is known about how and when immunocompromised hosts play a critical role in pathogen evolution. Here, we use a simple mathematical model to understand the effects of immunocompromised hosts on the emergence of immune escape variants in the presence and absence of epistasis. We show that when the pathogen does not have to cross a fitness valley for immune escape to occur (no epistasis), immunocompromised individuals have no qualitative effect on antigenic evolution (although they may accelerate immune escape if within-host evolutionary dynamics are faster in immunocompromised individuals). But if a fitness valley exists between immune escape variants at the between-host level (epistasis), then persistent infections of immunocompromised individuals allow mutations to accumulate, therefore facilitating rather than simply speeding up antigenic evolution. Our results suggest that better genomic surveillance of infected immunocompromised individuals and better global health equality, including improving access to vaccines and treatments for individuals who are immunocompromised (especially in lower- and middle-income countries), may be crucial to preventing the emergence of future immune escape variants of SARS-CoV-2.

The immune response to SARS-CoV-2 in people with HIV

“…It is thus clear that controlling HIV may contribute to controlling COVID-19. In addition to HIV treatment, modeling suggests that better antiviral treatments for SARS-CoV-2 infection for this group can substantially lower the probability of novel variants emerging [ 118 ].…”

Coordinated evolution at amino acid sites of SARS-CoV-2 spike

“…Finally, the signal of concordant evolution at the origin of VOCs could come from the combination of both factors: a distinct selection regime and epistasis. The role of immunocompromised patients in the evolution of SARS-CoV-2 is even higher for epistatic than for the non-epistatic mode of evolution ( Smith and Ashby, 2022 ). If the viral fitness, as it has been previously proposed ( Hill et al, 2022 ; Rochman et al, 2021a ; Smith and Ashby, 2022 ), is a trade-off between transmission efficiency and ability to avoid herd immunity, and the transmission component of the fitness landscape has valleys of low-fitness genotypes, the immunocompromised individuals with prolonged infectious periods due to relaxed selection for transmission efficiency allow the virus to accumulate mutations and cross these valleys.…”

“…The role of immunocompromised patients in the evolution of SARS-CoV-2 is even higher for epistatic than for the non-epistatic mode of evolution ( Smith and Ashby, 2022 ). If the viral fitness, as it has been previously proposed ( Hill et al, 2022 ; Rochman et al, 2021a ; Smith and Ashby, 2022 ), is a trade-off between transmission efficiency and ability to avoid herd immunity, and the transmission component of the fitness landscape has valleys of low-fitness genotypes, the immunocompromised individuals with prolonged infectious periods due to relaxed selection for transmission efficiency allow the virus to accumulate mutations and cross these valleys. Direct experimental studies of the possible epistatic interactions between coevolving sites will help elucidate the mechanism of origin of radically novel viral variants.…”

Non-pharmaceutical interventions and the emergence of pathogen variants

“…For example, several major variants of SARS-CoV-2 exhibit large numbers of mutations, especially in the spike protein [ 47 , 48 ]. If a large number of mutations are required to substantially escape host immunity, then epistasis may severely constrain pathogen evolution with immunocompromised hosts potentially crucial for antigenic evolution [ 50 ]. Fourth, we assumed that when NPIs were active, they were fixed at a constant level, which may be reasonable if they are in place for a relatively short period of time.…”

“…Population heterogeneity and contact structure both affect pathogen transmission, and so are also likely to affect the initial spread of variants, potentially allowing variants to gain a foothold in a subset of the population. Conversely, a more complex genetic and phenotypic landscape would likely make it more difficult for variants to emerge, for example, due to epistasis [ 50 ]. One could crudely model this by reducing the mutation supply in our model to mimic the lower rate of accumulating multiple mutations, which would quantitatively, but not qualitatively, change our results.…”